Verity processes videos by applying:
Computer Vision analysis to sampled image frames.
Natural Language Processing (NLP) analysis to text transcribed from the video’s audio track.
Verity runs parallel analysis of image and text components from the video. Verity samples frames from the video and sends them for image analysis. Verity extracts the audio track from the video and sends it for transcription from speech-to-text. Verity then enriches the text output with metadata (title, description) and sends it for NLP analysis.
The results of these parallel analyses are then merged in Verity’s proprietary merging logic. A comprehensive analysis considering both auditory and visual elements of the video is returned.